Vol. 21: 151-155, 1995

DISEASES OF AQUATIC ORGANISMS Dis. aquat. Org.

l

Published March 9

NOTE

Identification of Pseudomonas anguilliseptica isolated from several fish species in France Franck C. J. Berthe, Christian Michel, Jean-Franqois Bernardet* Unite d e Virologie et Immunologic Mol6culaires. Centre de Recherches INRA. F-78352 Jouy-en-Josas Cedex. France

ABSTRACT: Since 1990, outbreaks of haemorrhagic septicemia, frequently associated with keratitls, have occurred on gilthead sea bream Sparus aurata, farmed along the French Mediterranean and Atlantic coasts. Several bacterial isolates obtained from diseased sea bream, sea bass Dicentrarchus labrax, and turbot Scophthalmus maximus showed phenotypic similarities with the Pseudomonas anguilliseptica Mlakabayashi & Egusa, 1972 type strain NCIMB 194gT The French isolates were con~paredwith the type strain by determining levels of deoxyribonucleic acid (DNA) relatedness (S1 nuclease method), guanine + cytosine content (thermal denaturation method), and by performing phenotypic tests. All of the isolates formed a tight genomic species with higher than 90% DNA relatedness to the type strain. This study demonstrates the phenotypic and genomic homogeneity of P. anguilliseptlca and confirms that this bacterial species exhibits no host specificity. It thus represents a potential hazard for many farmed fish species and could cause serious economic losses. KEY WORDS: Pseudomonas anguillisept~ca. Sparus aurata . D~centrarchus labrax . Scophthalnlus maximus . Bacterial identification . DNA homology

Pseudornonas anguilliseptica was originally isolated by Wakabayashi & Egusa (1972) from pond cultured Japanese eels Anguilla japonica suffering from 'sekiten-byo' (red spot disease). This bacterial disease caused mass mortalities during early summer 1971 in Japan. Since then, P. anguilliseptica has been recorded from European eels Anguilla anguilla in Taiwan (Kuo & Kou 1978). In Europe, the pathogen was isolated in Scotland (Stewart et al. 1983) and later from Danish eel farms (Mellergaard & Dalsgaard 1986). The disease was also reported from other fish species: black sea bream Acanthopagrus schlegeli (Nakajima et al. 1983), ayu Plecoglossus altivelis (Nakai et al. 1985),salmonid fish (Wiklund & Bylund 1990) and more recently from wild Baltic herring Clupea harengus membras (Lonnstrom et al. 1994). The bacterium was first isolated in

'Addressee for correspondence O Inter-Research 1995

France by Michel and coworkers in 1992 from European eels reared in fresh water. Since 1990, outbreaks of haemorrhagic septicemia, frequently associated with keratitis, have occurred during winter in farmed gilthead sea bream Sparus aurata along the French Mediterranean and Atlantic coasts. The disease was only observed at water temperatures below 16°C and the highest mortality occurred at water temperatures in the 9 to 13OC range. Abdominal distension causing a so-called 'belly-up' syndrome was often observed. Petechial haemorrhages were noticeable on the skin and the liver but the lesions observed in these outbreaks were less severe and extensive in comparison with the histopathology described in the Japanese (Miyazaki & Egusa 1977) and European eel (Ellis et al. 1983). The bacterium was observed in blood smears and in some of the superficial dermis histological sections. Very recently, the disease was also noticed in farmed turbot Scophthalrnus maximus and sea bass Dicentrarchus labrax in France. Bacterial isolates obtained in pure culture from head-kidney and spleen appeared very similar in all of the outbreaks. Studies carried out to confirm that the isolates were Pseudornonas anguilliseptica are reported here. Materials and methods. The bacterial isolates included in this study are listed in Table 1 along with their source. The type strain of the species Pseudomonas anguilliseptica NCIMB 194gT was also included in all of the tests as a reference strain. The primary isolation of the bacteria was made on trypticase soy agar (TSA, bioMerieux, Charbonnieresles-Bains, France) at 22°C. Gram staining was performed on tissues or blood smears and checked again on purified cultures. The general procedures described by Le Minor & Veron (1989) and Carbonnelle & Kouyoumdjian (1987) were used at 22°C to perform the basic tests, unless otherwise stated.

152

Dis aquat. Org.

Table 1 P s e u d o n ~ o n a sangllilliseptlca lsolates included in thls study Isolate" NCIhiB 1949'

JIP 14/91

JIP 24/91 JIP 02/92-1

Source Blood, pond-cultured Japanese eel Anguilla]aponica, Shizuoka Prefecture, J a p a n , 1971 Head-kldney, gilthead sea bream Sparus aurata, M e d ~ t e r r a n e a ncoast, France, 1991 Head-kidney, gilthead sea bream, Medlterranean coast. France, 1991 Head-kidney, gilthead sea bream, Noirmoutiers Island, Atlantic coast, France, 1992

JIP 02/92-2

Same origln a s strain 02/92-1, but from another flsh

JIP 02/92-3

S a m e orlgin a s stralns 02/92-l a n d 2, but from another fish S a m e orlgln a s strains 02/92-l, 2 a n d 3 , but from another flsh

JIP 02/92-4 JIP 13/92-1 JIP 13/92-2

Head-kidney, gllthead sea bream, R e Island, Atlantlc coast, France, 1992 S a m e ongln as straln 13/92-l, but from another fish

JIP 13/92-4

Same origin a s strains 13/92-l a n d 2, but from another fish

LPAA 11066

Head-kldney, sea bass Dicentrarchus labrax, Brittany, Atlantic coast, France, 1991 Turbot Scophthalmus maximus, France, 1994

LPAA 11705 LPAA 11706

Turbot, France, 1994

LVDL 1291/94

Spleen, turbot, Spaln, 1.994

dNCIMB, Natlonal Collection of Industr~aland M a n n e Bacteria, Aberdeen, Scotland, Isolate provlded by Prof H Wakabayashl, Dept of Flshenes, Faculty of Agriculture, U n ~ v e r s ~ tofy Tokyo, J a p a n JIP, collect~onof isolates of the Unlte d e Vlrologle et Immunolog~eMoleculalres, Jouy-en-Josas, France LPAA, Laborato~re d e Pathologie des Anlmaux Aquatlques-Centre N a t ~ o n a l d'Etudes Vetennalres et Alimentalres, Brest, France, isolates provlded by Dr M V~gneulle,LVDL Laborato~re Vetennalre Departemental d e s Landes, Mont d e Marsan. France, lsolate provlded by Dr P Nougayrede

Motility was investigated in wet mounts and hanglng drops piepared from tryptlcase soy broth (TSB, bloMerieux) after lncubat~onat 15°C according to the recommendations of Wakabayash~& Egusa (1972) As growth of Pseudornonas anguillisept~ca on m a n n e Anacker 81 Ordal agar (mAOA) seems to enhance motility (H Wakabayash~pers comm ) , motillty was also checked on bacteria grown on mAOA slants AOA is composed of 0 05 % tryptone, 0 05 O/o yeast extract, 0 02% beef extract, and 0 02% sodlum acetate ( A n a c k e ~& Ordal 1955), mAOA is prepared cvlth 6 0 n r 1

seawater. The presence of flagella was checked by light microscopy after stalning as described by Kodaka et al. (1982), and by transmission electron microscopy: cells suspended In saline were placed on Formvarcoated grids, flxed, negatively stained with 1 % ammonlunl molybdate (pH 4.5),and observed with a Ph~lips EM12 transmission electron microscope operated at 80 kV The following tests were performed on the 13 French isolates a n d on the type strain using commercially available media (Diagnostics Pasteur, Marnes-laCoquette, France): respiratory metabolism; production of cytochrome oxidase, catalase, indole (Kovac's reagent), and urease; O/F test (modified Hugh & Leifson's medium, MEVAG); fermentation of lactose and glucose and production of H2S (Kliger's medium); reduction of nitrate (1 % nitrate broth); utilization of sodium citrate (Simmons' medium); tolerance to temperature (5, 30, and 37°C) and salinity (0, 1, and 3 % NaC1) in l % peptone broth; o-nitrophenyl-P-D-galactopyranoside test; degradation of aesculin, Tweens 20 and 80 (according to Sierra 1957), DNA, casein (10% skim milk TSA), and starch (0.5 % soluble starch TSA). Gelatin degradation was tested for by the film method (Le Minor & Piechaud 1963): a strip of exposed photographic film was partially immersed In a thick bacterial suspension in saline, and the possible liquefaction of the thin black gelatin layer was observed after a 4 d incubation. Enzymatic profiles were determined in API ZYM galleries (bioMerieux) incubated overnight at 22°C. The Pseudornonas anguilliseptica type straln NCIMB 194gTwas used to prepare a specific rabbit antiserum: saline suspensions containing increasing concentrations of formalin-killed bacterial cells were injected intravenous1.y. Live cells of all of the isolates were tested in a rapid slide agglutination test against the non-diluted anti-Pseudornonas anguilliseptica serum. The agglutinin titer o[ the serum against homologous antigen was 1/128. Previously described methods were used to extract, purify, and shear unlabeled DNAs (Brenner et al. 1982).The procedure used for in vitro labeling of DNA with a tritium-labeled nucleotide ([1',2',5'-3H]dCTP, Amersham, TRK625) has been published by Grimont et a1 (1980), and modified by using a Megaprlme kit (Megaprime DNA labeling system, Amersham, RPN1606). The method for hybridization experiments (S1 nuclease-DE81 method) has been described by Popoff et al. (1981). The guanine-plus-cytosine content (G+C:b) of the bacterial DNAs was determined by the thermal denaturation method and calculated using the equation of Marmur & Doty (1962) as modified by De Ley (1970).

Berthe et al.. Pseudornonas ang~~ilhsepfica from fish in France

Results and discussion. In pure culture, the isolates examined exhibited slow growth on TSA, producing tiny translucent non-pigmented colonies. The results of the phenotypic tests performed on the 13 French isolates are shown in Table 2, together with the characteristics of the type strain NCIh4B 194gT.From this limited range of classical phenotypical tests, the French isolates proved very similar Because of their rod shape, Gram-negative staining, polar flagellation, strict aerobic respiration, and production of cytochrome oxidase, these isolates were placed in the genus Pseudomonas. On the basis of morphological, physiological, and biochemical characteristics they were tentatively identified as P. anguilliseptica. In the course of this study, motility was not observed in hanging drops and wet mounts prepared from cultures in TSB. However, it was possible to observe motility with some of the isolates grown on mAOA slants; flagella were observed by light microscopic Table 2. Phenotypic characteristics of 13 French Pseudonlonas anguilliseptica isolates and of the type strain NCIMB 1 9 4 9 ~ French isolates Gram stain Morphology Motility Cytochrome oxidase Catalase O/F test H,S production Nitrate reduction Citrate Simmons Fermentation of: Gl~~cose Lactose Growth at: 5°C 30°C 37°C

Growth in: 0% NaCl 1 NaCl 3%)NaCl ONPG test Indole production Urease production Degradation of: Casein DNA Aesculin Gelatin Starch Tween 20 Tween 80

NCIMB 1949'

-

rod

rod

+

+

+

+

+

t

+

-

-

-

-

11/13"

-

+ -

+ +

+ +

-

t t t

" l l of the 13 isolates studied gave a positive result

153

examination and by electron microscopy. Wakabayashi & Egusa (1972) observed 2 types of cells by electron microscopy: one had a flagell.un1, while the second had no flagellum but a coiled flagellurn-like structure within the envelope of the cell. From these observations, they concluded that Pseudomonas anguilliseptica exhibits a temperature-dependent expression of motility. The variability of motility with temperature was confirmed by the fact that motile cells increased in number when the culture, initially grown in a 25°C incubator, was moved to 15°C. Variable motility was also reported by Shea et al. (1991) during a study of the gliding bacterium Deleya marina: a subpopulation of swimming cells appeared randomly and reversibly on subculture, exhibiting a surface-associated motility. Moreover, motility of Flavobacterlum aquatile, the type species of a non-motile genus, was discussed by Holmes & Owen (1979),highlighting the possible variability of this characteristic. Thus, in reporting a culture as non-motile, the test conditions must be precisely specified and the culture should be shown to lack flagella. Very few phenotypic discrepancies occurred among the isolates tested in the course of this study. The ability to grow in a 0% NaCl peptone medium was exhibited by all of the isolates. This result was consistent with the data available for the type strain NCIMB 194gTfrom Wiklund & Bylund (1990) and Lonnstrom et al. (1994). However, in the original description of the species (Wakabayashi & Egusa 1972),no growth in 0 % NaCl was recorded. Salt tolerance is usually determined using a nutrient broth supplemented with 0.1 %, MgC12, and containing different NaCl concentrations (Wiklund & Bylund 1990). The requirement for elements such as Mg", Ca" or K+ by many bacteria isolated from the marine environment has been reported (Bertrand & Larsen 1989). Different concentrations of these elements in the media used for testing the ability to grow without NaCl could thus explain the discrepancies between data from different authors. However, Lonnstrom et al. (1994) showed that Pseudornonas anguilliseptica was capable of growing on Anacker & Ordal's medium, which contalns relatively low amounts of nutrients and which was not supplemented with NaCl. The French isolates and the type strain proved unable to hydrolyze gelatin with the film method. Most Pseudomonas anguilliseptica isolates from salmonid fish studied by Wiklund & Bylund (1990) also failed to hydrolyze gelatin (method used unspecified). However, these authors as well as Wakabayashi & Egusa (1972) found the type strain NCIMB 194gT to be positive for this test. Michel et al. (1992) observed that P. anguilliseptica strains initially capable of hydrolyzing gelatin readily lost this property after subculture.

154

Dis. aquat. Org. 21: 151-155, 1995

available from Nakai et al. (1981), some Listonella angulllara (Vibrio anguillarum) and Aeromonas salmonicida s t r a ~ n swere checked with this antiserum for posslble cross reaction: no NC[hlB 13 French Hydrolysis of the 194%' isolates follo\ving substrates cross reaction was observed. The base composition of the DNA of 7 repre+" 2-Naphthyl-phosphate sentative isolates (JIP 02/92-l to 4 and 13/92-1, 2, 2-Naphthyl-butyrate and 4) and of the type strain NCIMB 1 9 4 9 was ~ in t 2-Naphthyl-caprylate . . 2-~a~hth~l-m;ristate + the 59 to 61 mol'!:, range as previously deterL-leucyl-2-naphthglamide + + mined for the type strain a n d 2 Pseudornonas L-valyl-2-naphthylamide anguilliseptica isolates from eels using high L-cystyl-2-naphthylamide pressure liquid chromatography (Michel et al. N-benzoyl-DL-arginine-2-naphthylamide -9/13~ N-glutaryl-phenylalanine-2-naphthylamide 1992). The lower G+C content published by + + 2-Naphthyl-phosphate Stewart et al. (1983) has been discussed elseNaphthol-AS-BI-phosphate + + where (Michel et al. 1992). 6-Br-2-naphtyl-aD-galactopyranoside The DNA-DNA hybridization studies showed 2-Naphthyl-PD-galactopyranoside that the 7 selected French isolates were more Naphthol-AS-BI-DD-glucuronide 2-Naphthyl-aD-glucopyranoside than 9 0 % related to the type strain NCIMB 6-Br-2-naphtyl-BD-glucopyranoside 1 9 4 9 ~With . the method used in this study, 2 bacl-Naphthyl-N-acetyl-D-glucosaminide terial isolates are considered to belong to the 6-Br-2-naphthyl-aD-mannopyranoside same species when their DNA relatedness is at 2-Naphthyl-aL-fucopyranoslde least 7 0 % . A more robust definition (80% relat"API reaction scores 0 and 1 were considered negative; scores 2 ed.ness) was given by Grimont (1988); because to 5 were considered positive our isolates complied with this definition, w e felt "9 of the 13 isolates tested gave a negative result no need for studying the thermal stability of the DNA hybrids. Our phenotypic and genornlc investigations Another discrepancy should be pointed out in thus demonstrated that the French isolates belong to the species Pseudomonas anguilliseptica. This bacterTable 2: DNAse production was found to be positive for the French isolates a n d the type strain, but was ial species a.ppears very homogeneous as no characterreported to be negative by Lonnstrijrn et al. (1994). As istics capable of distinguishing the French stralns from the method they used was not specified, it is possible the type strain NCIMB 1949T were found. To investigate the pathogenicity of the isolates, some that t h e discrepancy was d u e to the use of different methodological approaches. preliminary experimental infection tests were performed on several 10-fish groups of 40 g sea bream T h e enzymatic profiles in API ZYM galleries a r e pre(J.-F. Pepin pers. cornrn.). Bacterial suspensions at difsented in Table 3. T h e profiles of the 13 French isolates were found to be very similar, except for the N-benferent concentrations (1x107 to 4.8 x107 CFU fish-') in zoyl-DL-arginine-2-naphthylamide test. The only dissaline were inoculated by 3 routes: intraperitoneal crepancy with the profile of the type strain occurred injection, subcutaneous injection, a n d swabbing after with 2-naphthyl-myristate: this substrate was hydroscarification. All of the challenged fish died within a lyzed by the type strain but by none of the French isofew days without exhibiting any external signs of dislates. ease a n d the bacterium was reisolated from all of the The slide agglu.tination reaction with the antiserum dead fish. Jn contrast, unchallenged controls all surraised again.st the Pseudomonas anguilliseptica type vived. strain was positive with all of the French isolates. An The Pseudomonas anguilliseptica infections are a antiserum was also raised against a French P. anguilcause for concern because the mortality is persistent liseptica isol.ate from sea bream not included in this and treatment is only partially effective. The bacterium was isolated from several farmed fish species and study (M. Vigneulle pers. comm.); the slide agglutination reaction was positive with all of the French isolates exhibits no host specifici.ty. Since its first identification studied here, while the type strain did not agglutinate in Japan 22 yr ago, P. anguilliseptica has been found to occur in Scotland (Stewart et a.1. 1983), Finland (Wikwith this antiserum. As 2 different serotypes of P. ang~lillisepticahave been observed in Japan (Nakai et lund & Bylund 1990), Denmark (Mellergaard & Dalsal. 1981), it is possible that the French isolate used for gaard 1986), and France (Michel et al. 1992). Large portlons of the European coast seem to be affected; preparing the antiserum was of a. different serotype than the type strain. Taking into account other data thus, this fish pathogen should be a growing source of Table 3. API 2Yx.l prof~lesof the 13 French isolates and of the Pse~idomonasangullliseptica type strain NCIMB 1949'

Bel-the et a1 : Pseudornonas angurlhseptjca fro111flsh In France

concern for marine aquaculture in Europe. Moreovel-, because this widespread pathogen was recently isolated from wild fish (Lonnstrom et al. 1994), there is a need for further epizootiologic studies on P. angullliseptica.

155

Anacker RL, Ordal EJ (1955) Study of bacteriophage infecting the myxobacterium Chondrococcus c o l u m ~ ~ a nJs . Bacteriol 70:738-741 Bertrand J C , Larsen H (1989) La 'bacterie marine': mythe ou realite In: Bianchi M, h4arty D, Bertrand J C , Caumette P, Gauthier M (eds) Micro-organismes dans les ecosystemes oceaniques. Masson, Paris, p 3-25 Brenner DJ, McWhorter AC, Leete Knutson JK. Steigerwalt AG (1982) Escherichia vulnei-1s. a new species of Enterobacteriaceae associated with human wounds. J c l ~ nMicrobiol 15 1133-1 140 Carbonnelle B, Kouyoumdjian S (1987) Les techniques et les etapes d e l'analyse. In. Carbonnelle B, Denis F, Marinonier A, Pinon G , Vargues R (eds) Bacteriologie medicale SIMEP, Pans, p 9-30 De Ley, J (1970) Reexamination of the associat~onbetween melting point, buoyant density, and chemical base composition of deoxyribonucleic acid. J Bacteriol 101.738-754 E l l ~ s AE, Dear G, Stewart, DJ (1983) Histopathology of 'sekiten-byo' caused by Pseudornonas angu~llisepticain the European eel, Angujlla angullla L, In Scotland. J Fish Dis 6:77-79 Grimont PAD (1988) Use of DNA reassociation in bacterial classification. C a n J Microbiol 34:541-545 Grimont PAD, Popoff MY, Gi-imont F, Coynault C , Lemelin M (1980) Reproducibility a n d correlation study of three deoxyribonucleic acid hybridization procedures Curr Microbiol 4:325-330 Holmes B, Owen RJ (1979) Proposal that Flavobacterium breve be substituted as the type specles of the genus in place of Flavobacterium aquatlle and emended description of the genus Flavobacteriurn: status of the named species of Flavobactenum Int J system Bactenol 29:416-426 Kodaka H, Armfield AY, Lombard GL, Dowel1 VR (1982)

Pract~calprocedure for demonstrating b a c t e r ~ a lflagella. J clin Microbiol 16-948-952 Kuo S - C , Kou G - H (1978) Pseudoinonas a n g u r l l ~ s e p t ~ c)soa lated from red spot disease of pond-cultured eel, Angililla japonjca Rep Inst Fish Biol, Mln Econ Aff, Nat Tailvan Univ 3:19-23 Le Minor L, Piechaud M (1963) Uric, methode rapide d e recherche d e la proteolyse d e la qrlatine Annls lnst Pasteur, Paris 105 792-794 Le h4inor L, Veron M (1989) Bactenologie medlcale. h4edecines-Sciences Flammarion, Paris Lonnstrom L, Wiklund T, Bylund G (1994) Pseildomonas a n g ~ ~ i l l l s e p t ~isolated ca fi-om Baltic herring Clupea hareng u s nlernbras with eye leslons. Dis aquat Oi-g 18.143-147 Marmur J , Doty P (1962) Determination of the base composttion of deoxynbonucleic acid from its thermal denaturat ~ o ntemperature. J molec Biol 5.109-118 Mellergaai-d S, Dalsgaard I (1986) HBndbog i Alesyydomme. Danm Fisk- og Havunders Rapp 293:33-34 Michel C, Bernardet J-F, Dinand D (1992) Phenotypic a n d genotypic studles of Pseudomonas angudl~septicastrains isolated from farmed European eels (Angu~llaa n y ~ l i l l ain ) France. Fish Pathol 27:229-232 Miyazaki T, Egusa S (1977) Histopatholog~calstudies of red spot disease of the J a p a n e s e eel (Anyuilla japonlca) I . Natural infection. Fish Pathol 12:39-49 Nakal T, Hanada H , Muroga K (1985) First records of Pseudornonas a n g u ~ l l ~ s e p t ~~nfection ca In cultured ayu, Plecoglossus altjveljs Fish Pathol 20 481-484 Nakai T, Muroga K, LVakabayashi H (1981) Serolog~calproperties of Pseudornonas a n g ~ ~ ~ l l i s e p t in i c aagglutination Bull J a p Soc scient Fish 47:699-703 Nakajima K , Muroga K , Hancock RECV (1983) Comparison of fatty acid, protein, and serological properties distinguishing outer membranes of Pseudornonas a n g u ~ l l j s e p t ~ c a strains fi-om those of fish pathogens and other pseudomonads. lnt J system Bacteriol 33:l-8 Popoff MY, Coynault C , Kiredjian M, Lemelin M (1981) Polynucleotide sequence relatedness among motlle Aeromonas species. Curr Microbiol 5:109-114 Shea C , Nunley JW, Smith-Somerville HE (1991) Variable e x p ~ e s s ~ oofn gliding and swimmlng motility in Deleya m a n n a Can J Microbiol 37 808-814 Sierra G (1957) A simple method for the detection of lipolytic activity of microorganisms and some observations of the influence of the contact between cells a n d fatty substrates. A n t o n ~ evan Leeuwenhoek 23.15 Stewart DJ, Woldemanam K, Dear G , Mochaba FM (1983) An outbreak of 'sekiten-byo' among cultured European eels, Angullla angullla L, in Scotland. J Fish Dis 6:75-76 Wakabayashl H, Egusa S (1972) Characterist~csof a Pseudomonas sp. from a n epizootlc of pond-cultured eels (Anguilla japonica). Bull J a p Soc scient Fish 38 577-587 Wiklund T, Bylund G (1990) Pseudomonas anguilliseptica a s a pathogen of salmonid fish in Finland. Dis aquat Org 8:13-19

Responsible Subject Editor- T. Evelyn, Nanalmo, B.C., Canada

Manuscript first received: J u n e 20, 1994 R e v ~ s e dversion accepted: October 25, 1994

Acknoruledgernents. We a r e grateful to Dr M Vigneulle (LPAA-CNEVA, Brest. France) for supplying the antiserum against the French isolate; \Ire a r e ~ n d e b t e d to Dr M. Vigneulle, PI-of.H. Wakabayashi (Dept of Fisheries, Faculty of Agriculture, Un~versity of Tokyo, J a p a n ) , and Dr P Nougayrede (LVD, Mont d e Marsan, France) for kindly providing some of the strains used in this study. to R. Tytgat (Laboratorium voor Microbiologie, Uni\iersiteit Gent, Belgium) for determining the base composition of the DNAs, and to Dr S. Chilmonczyk a n d B Flechon (INRA, Jouy-en-Josas, France) for kind help with the electron mlcroscopy.

LITERATURE CITED